position of subject in nonfinite clauses

[J. Clancy Clements (Kapil)]

On Tue, 17 Feb 1998, Diego Quesada wrote:

>         My black-box agrees with Jon's assessment of the preinfinitival
> subject. I hear things like this all the time
>
>         (1)     Por vos andar de hocicon ahora...
>
>         (2)     Por Juan estar de guardia ese dia...
>
> In the Dominican Republic and other Caribbean places, one hears
>
>         (3)     Para yo saber que tu dices
>
>         (4)     Para tu decirme si esta bien.

Caribbean Spanish allows this.  Correct.  Castilian Spanish allows it
marginally.  For example, Diego's (2) and (4) are bad in CS, whereas (3)
is somewhat better.  The following varieties allow it marginally: Mexican
(non-Caribbean), Argentinan, Chilean.  I'm not sure about Peruvian,
Bolivian, etc.  The question can now be recast:  what functional
explanation can there be for the more unrestricted presence of it in some
dialects, and the marginal presence of it in other dialects?  Note: the
presence of it in CS and other varieties seems to have to do with the
realization of the subject: if it's a full NP, it's worse, if it's a
pronoun, it's not as bad. This is a GENERAL characterization, which for a
functional analysis is not helpful because it's not possible to see how
things function in a general characterization of a phenomenon.  Still, if
anyone has an idea, functionally, why some varieties tend to disallow it
and other varieties allow it, I really like to hear it.

Thanks in advance (and to Brian M. for his response).

Clancy